Discussion Thread for the Continuous GW Search known as O2MD1 (now O2MDF - GPUs only)

Mr P Hucker
Mr P Hucker
Joined: 12 Aug 06
Posts: 819
Credit: 481369262
RAC: 4124

Ian&Steve C. wrote: I've

Ian&Steve C. wrote:

I've posted the relevant information in the tech forum, but I can't make the required people read it, and/or they may have other priorities. 

theres a better chance of it getting more attention if more people bring it up, not just one man.

the squeaky wheel gets the grease. or something like that.

I could start a petition on care2?

If it's causing them a problem, I'm sure they'll notice it from their end.  According to the server status, the majority are completing ok.

If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.

Mr P Hucker
Mr P Hucker
Joined: 12 Aug 06
Posts: 819
Credit: 481369262
RAC: 4124

Betreger wrote:I am

Betreger wrote:

I am starting to think they are willing to accept the overhead. In my case I was still producing good results 2/3 of the time. 

As an aside my 3GB cards are now only doing pulsars, the 6GB card get GWs. 

I gave up on GW altogether, not due to RAM, but due to my CPUs being rubbish at assisting.  I did have one combo which worked, but it requires running 2 GW on the GPU so the CPU can allocate 2 cores to help.  But now they're too big and go very slowly.  I build PCs from cheap parts and they run what they can run.  I like recycling.

If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.

Holmis
Joined: 4 Jan 05
Posts: 1118
Credit: 1055935564
RAC: 0

Peter Hucker wrote: tullio

Peter Hucker wrote:

tullio wrote:

Watching GW GPU tasks on my BOINC manager I see a curious thing. Progress rises very rapidly up about 14% in about 3 minutes, then falls back to 0.470 and rises more slowly.

Tullio

That happens on many projects, I wouldn't worry about it.  It's just the task going through two different stages and having a rubbish progress meter.

Actually it's Boinc that's impatient and starts simulating progress if the science application hasn't reported any progress after 1 minute. Then when the app starts reporting progress done Boinc switches over to the real progress reported by the app.

Mr P Hucker
Mr P Hucker
Joined: 12 Aug 06
Posts: 819
Credit: 481369262
RAC: 4124

Holmis wrote: Peter Hucker

Holmis wrote:

Peter Hucker wrote:

That happens on many projects, I wouldn't worry about it.  It's just the task going through two different stages and having a rubbish progress meter.

Actually it's Boinc that's impatient and starts simulating progress if the science application hasn't reported any progress after 1 minute. Then when the app starts reporting progress done Boinc switches over to the real progress reported by the app.

Ah that makes sense, the initial count does look rather like a timer rather than actual progress, which would be less even.

Although, when I load several Theory tasks from LHC at once onto my 24 core computer without an SSD, it can take ages for them to start as the disk tries to catch up, and no counter appears.  Perhaps Boinc realises the disk is busy and doesn't get so impatient?

If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.

Holmis
Joined: 4 Jan 05
Posts: 1118
Credit: 1055935564
RAC: 0

I did a quick google to try

I did a quick google to try and find the documentation for this "feature" but couldn't find it :(

It's quite possible that the Boinc API allows for the app to tell Boinc that it's still loading stuff and hasn't begun processing yet. As I'm no coder/developer so I just don't know.

Keith Myers
Keith Myers
Joined: 11 Feb 11
Posts: 4704
Credit: 17548586759
RAC: 6436592

There is "pseudo-progess"

There is "pseudo-progess" code in the client for those projects that don't report % completed.

You can typically see the progress percentage climb quickly in small incremental jumps within the first minute or so of starting a task until the task writes out a checkpoint and then reports its actual progress.

DA put that piece of code in for some of the very long completion tasks on some projects. Trying to reduce the number of posts/complaints that people never see any task move towards a finish and they think the task is stalled out.

 

Mr P Hucker
Mr P Hucker
Joined: 12 Aug 06
Posts: 819
Credit: 481369262
RAC: 4124

Keith Myers wrote: There is

Keith Myers wrote:

There is "pseudo-progess" code in the client for those projects that don't report % completed.

You can typically see the progress percentage climb quickly in small incremental jumps within the first minute or so of starting a task until the task writes out a checkpoint and then reports its actual progress.

DA put that piece of code in for some of the very long completion tasks on some projects. Trying to reduce the number of posts/complaints that people never see any task move towards a finish and they think the task is stalled out.

Doesn't that mean if the task has stalled, it looks like it's running ok?

If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.

Keith Myers
Keith Myers
Joined: 11 Feb 11
Posts: 4704
Credit: 17548586759
RAC: 6436592

Don't know.  Probably not as

Don't know.  Probably not as there should be code that detects an actual stall.  I'd have to find the code again in the client where it got implemented.

Richard Haslegrove would know for sure and can probably find the code blob much quicker than I could.

 

Tom M
Tom M
Joined: 2 Feb 06
Posts: 5589
Credit: 7675565996
RAC: 1845797

Linux Radeon VII boxIt

Linux Radeon VII box

It looks like you can run at least 4 GW gpu ati-beta tasks on a linux-based system with a Radeon VII gpu.

For some reason I could not get it to run more than 4 of the current ati tasks. And six beta maybe unstable. 7 is unstable.

I also had trouble trying to run a mix of Gamma Ray and GW when I used different thread counts.  I have not tested 6 Gamma Rays (Pulsar search#1) yet.  Since my goal was maximum GW production on one card/one system.

Tom M

A Proud member of the O.F.A.  (Old Farts Association).  Be well, do good work, and keep in touch.® (Garrison Keillor)

Ian&Steve C.
Ian&Steve C.
Joined: 19 Jan 20
Posts: 3681
Credit: 33842006021
RAC: 36991912

is there a reason you keep

is there a reason you keep wanting to run the beta app? seriously. is it faster? or was it an arbitrary decision?

_________________________________________________________________________

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.